This study examined factors that influence anchorage loss (AL) during space closure following maxillary premolar extractions. 87 patients were divided into groups based on extraction site, appliance type, and age. AL was measured on lateral cephalograms and dental casts. The results showed that increased crowding was associated with decreased AL. Labial appliances showed greater AL than lingual appliances. Adults tended to have greater AL than adolescents, though not significantly. Extraction site and overjet were weakly correlated with AL. The primary factors influencing AL were crowding and appliance type, while age, extraction site, and overjet were secondary factors.

1. 730
Angle Orthodontist, Vol 73, No 6, 2003
Original Article
Anchorage Loss—A Multifactorial Response
Silvia Geron, DMD, MSca
; Nir Shpack, DMD, MScb
; Samouil Kandos, DMDc
;
Moshe Davidovitch, DMDa
; Alexander D. Vardimon, DMDd
Abstract: Anchorage loss (AL) is a potential side effect of orthodontic mechanotherapy. In the present
study, it is defined as the amount of mesial movement of the upper first permanent molar during premolar
extraction space closure. In addition, AL is described as a multifactorial response in relation to the ex-
traction site, appliance type, age, crowding, and overjet. For this study, 87 university clinic and private
practice subjects, who were defined as maximum anchorage cases and had undergone bilateral maxillary
premolar extractions, were divided into four groups according to extraction site (first vs second premolars),
mechanics (lingual vs labial edgewise appliances), and age (adolescents vs adults). Overjet and crowding
were examined from the overall sample. Data were collected from serial lateral cephalograms and dental
casts. The results showed that as the severity of dental crowding increased, AL significantly decreased (r
5 20.66, P 5 .001). Labial edgewise appliances demonstrated a significantly greater AL than did lingual
edgewise appliances (1.15 6 2.06 mm, P , .05). A greater, though not statistically significant, AL was
found in adults than in adolescents (0.73 6 1.43 mm). There was a slight nonsignificant increase in AL
between maxillary second compared with first premolar extractions (0.51 6 1.33 mm). Overjet was weakly
correlated to AL. These results suggest that AL is a multifactorial response and that the five examined
factors can be divided into primary (crowding, mechanics) and secondary factors (age, extraction site,
overjet), in declining order of importance. (Angle Orthod 2003;73:730–737.)
Key Words: Anchorage; Lingual orthodontics; Age; Extraction
INTRODUCTION
Anchorage is the resistance to unwanted tooth move-
ment1,2
and is commonly described as the desired reaction
of posterior teeth to space closure mechanotherapy to
achieve treatment goals, ie, minimum, medium, maximum
anchorage.3
Anchorage loss (AL) is a reciprocal reaction
that could obstruct the success of orthodontic treatment by
complicating the anteroposterior correction of the maloc-
clusion and possibly detracting from facial esthetics. A ma-
jor concern when correcting severe crowding, excessive
overjet, and bimaxillary protrusion is control of AL. There-
a
Clinical Instructor, Department of Orthodontics, The Maurice and
Gabriela Goldschleger School of Dental Medicine, Tel Aviv Univer-
sity, Tel Aviv, Israel.
b
Senior Clinical Instructor, Department of Orthodontics, The Mau-
rice and Gabriela Goldschleger School of Dental Medicine, Tel Aviv
University, Tel Aviv, Israel.
c
Private practice, Thessalonki, Greece.
d
Chairman, Department of Orthodontics, The Maurice and Gabrie-
la Goldschleger School of Dental Medicine, Tel Aviv University, Tel
Aviv, Israel.
Corresponding author: Silvia Geron, DMD, MSc, Orthodontic De-
partment, Tel Aviv University, Tel Aviv, Israel 69978
(e-mail: jgeron@zahav.net.il).
Accepted: February 2003. Submitted: October 2002.
q 2003 by The EH Angle Education and Research Foundation, Inc.
fore, adjunct appliances, such as the Nance holding arch,
transpalatal bar, and extraoral traction, are often used to
augment molar anchorage. The use of multiple teeth at the
anchorage segment to form a large counterbalancing unit
and the application of differential moments have also been
described as methods to stabilize molar position.3–5
Factors such as malocclusion, type and extent of tooth
movement (bodily/tipping), root angulation and length,
missing teeth, intraoral/extraoral mechanics, patient com-
pliance, crowding, overjet, extraction site, alveolar bone
contour, interarch interdigitation, skeletal pattern, third mo-
lars, and pathology (ie, ankylosis, periodontitis) affect AL.
Most of the AL studies focus on biomechanical solutions.4–10
For example, a greater AL was demonstrated with nonslid-
ing mechanics when the Gjessing spring was evaluated.6
Tip-edge brackets showed less but not significant AL than
straight wire brackets (1.71 vs 2.33 mm).7
Differential mo-
ments have been reported to reduce AL by 0.6–0.7 mm.4,5
When maximum anchorage is required, AL was greater in
Class I (0.60 mm) than in Class II (0.28 mm) malocclu-
sions.4
AL has also been referred to as the extent of incisor
protraction (1.8–2 mm) following molar distalization with
repelling magnets or nickel titanium open coil springs8,9
supported by a Nance appliance or the second premolar (1.6
mm). However, minimal incisor anchorage loss (0.92) has

2. 731
ANCHORAGE LOSS—A MULTIFACTORIAL RESPONSE
Angle Orthodontist, Vol 73, No 6, 2003
been reported when the pendulum appliance was used.10
Molar AL had occurred supported with an implant in the
center of the anterior palate (0.7–1.1 mm), but this was
probably caused by deformation of the transpalatal bars that
linked the implant to the maxillary molars.11
The concept of a well-interdigitated occlusion acting to
enhance molar anchorage is an accepted dogma. Therefore,
it could be hypothesized that the posterior disocclusion
caused by the anterior bite plane effect of a lingual appli-
ance might negate this. Thus, the decision to extract is more
frequent when lingual brackets are applied in the maxillary
arch.12
Lingual archwires are more rigid because of the
smaller interbracket distance.13
Extraction site is another factor that affects AL. Studies
conducted on the effect of the Begg appliance show that
the maxillary molar occupies 33.5% of the extraction site
with first premolar extractions and 50.4% with first molar
extractions.14
Creekmore15
found that the posterior teeth oc-
cupy one-third to one-half of the extraction space in first
and second premolar extractions, respectively. Furthermore,
in another study,16
no significant difference in AL was
found between first or second maxillary premolar extrac-
tions (4.3 vs 4.5 mm). However, when maxillary first pre-
molars were extracted in conjunction with mandibular first
or second premolars, AL of the maxillary molars was great-
er when the mandibular second premolars were extracted
(3.7 vs 4.7 mm).16
Dental crowding and its relationship to AL provide the
first sign that it is a multifactorial response. Second pre-
molar extraction, rather than first, is carried out far more
often in cases with less crowding. This choice has been
related to greater molar mesial movement.17
Additionally,
the maxillary chordal arch length (distance from mesial
contact point of the first molar to the contact point of the
central incisors) was reported to decrease in extraction cas-
es by 11.3 mm according to Ong and Woods16
and by 8.3
mm as reported by Luppanapornlarp and Johnston.18
This
difference corresponds to greater crowding found in the lat-
ter (5.8 mm) than in the former study (3.5 mm).
The effect of patient age on AL has not been widely
reported. Growing patients (12.5 years) experience 2.52
mm of AL, whereas nongrowing patients (27.6 years) show
an anchorage gain of 0.20 mm. The molar relationship is
corrected by mandibular growth in the adolescent group
(70%) and by maintaining maxillary molar position in the
adult group.19
It would appear that AL is seemingly dependent on more
than one factor, which, up to now, has been investigated
separately. The objectives of this study were to examine the
contribution of five such factors: extraction site (first vs
second premolars), mechanics (lingual vs labial technique),
age (growing vs nongrowing patients), crowding, and over-
jet and to determine their relative contributions to AL (pri-
mary vs secondary AL factors).
MATERIALS AND METHODS
The files of 211 subjects treated in the Tel Aviv Univer-
sity’s clinic and private practices were examined for ade-
quacy of diagnostic records. From these, a study sample
consisting of 87 Class I and Class II subjects was found.
Subjects were treated with either lingual (n 5 25) or labial
edgewise appliances (n 5 62). All lingually treated subjects
were nongrowing, whereas 20 of the labially treated sub-
jects were nongrowing as determined by chronological age
and normal growth curves. All subjects had undergone or-
thodontic treatment, which included extraction of two max-
illary first or second premolars. If only maxillary premolars
were extracted, treatment goals included Class I canine re-
lationship and Class II molar occlusion. Seven subjects
from the lingual group also had mandibular second pre-
molars extracted, making their treatment goals include both
a Class I canine and molar relationship.
All subjects were defined as maximum anchorage cases,
requiring minimal or no anterior movement of the molars
during space closure. Patients were included if the sum of
their maxillary dental crowding plus double the amount of
overjet together was greater than or equal to 11 mm
[crowding 1 (2 3 overjet) .11 mm],
ie, overjet varied from two to 13 mm and crowding varied
from one to 10 mm.
The sample was divided into four groups.
• G1—Nongrowing subjects treated with maxillary first
premolar extractions and lingual appliances (n 5 12, age
24.8 6 5.57 years).
• G2—Nongrowing subjects treated with maxillary second
premolar extractions and lingual appliances (n 5 13, age
24.4 6 5.99 years).
• G3—Nongrowing subjects treated with maxillary first
premolar extractions and labial appliances (n 5 20, age
20.09 6 5.43 years).
• G4—Growing subjects treated with maxillary first pre-
molar extractions and labial appliances (n 5 42, mean
age 12.6 6 1.99 years).
Comparisons were made such that when one factor (ex-
traction site, mechanics, or age) was examined using a
paired group comparison, the other two factors were non-
variable. Crowding and overjet were evaluated from the
paired-groups and the overall sample.
The labial subjects were treated with 0.022 3 0.028-inch
preadjusted brackets (Victory System, 3M Unitek, Monro-
via, CA), according to a standardized maximum anchorage
control regimen. The regimen included space closure by
individual (sliding) canine retraction followed by en masse
incisor retraction carried out on a 0.017 3 0.025-inch stain-
less steel archwire containing Bull-loops activated one mm
every four weeks and producing an initial force of 150 g
per side. Archwires were preactivated with tip-back bends

3. 732 GERON, SHPACK, KANDOS, DAVIDOVITCH, VARDIMON
Angle Orthodontist, Vol 73, No 6, 2003
FIGURE 1. The distance of the distal contact point of the maxillary
first molar to the occlusal line perpendicular (OLp) was measured
from pre- and posttreatment cephalometric radiographs, and the dif-
ference (post 2 pre) was defined as anchorage loss.
immediately distal to the premolar bracket. All subjects
treated with labial appliances were instructed to wear a cer-
vical headgear, containing an elevated long outer bow, 12
hours daily. For the remainder of the day, subjects were
instructed to wear Class II elastics (3/16-inch medium de-
livering 70–100 g).20
Compliance was verified by clinical
observation of molar position.
All lingual subjects were treated with bidimensional21
or-
thodontics using preadjusted brackets with a 0.018 3
0.025-inch archwire slot from canine-to-canine and an
0.022 3 0.028-inch slot in the posterior dentition (Ormco
Corp, Glendora, CA). These brackets were oriented and
bonded using a lingual jig method.22
Treatment was admin-
istered according to a standardized maximum anchorage
control regimen in which the second molars were included
in all cases and no headgear was used. Space closure was
accomplished by en masse sliding retraction of the anterior
teeth using intraarch orthodontic elastomeric chains, pro-
ducing initial forces of 150–200 g per side, approximately
50 g for each tooth. The elastic chain was replaced every
six weeks. All subjects treated with lingual appliances were
instructed to wear Class II elastics (3/16- or 1/8-inch me-
dium delivering 70–100 g) for eight hours a day.20
Space
closure was accomplished using a 0.016 3 0.022-inch
stainless steel archwires preactivated with compensating
curves (exaggerated Curve of Spee in the maxillary arch
wire and a reverse curve in the mandibular arch wire). Lin-
gual space closure was carried out by en masse retraction
of the six anterior teeth to avoid opening unaesthetic an-
terior spaces during the canine retraction phase.
In the lingual cases, the decision to extract first or second
premolars was based on the severity of anterior crowding.
Anterior crowding greater than six mm inclined toward first
premolar extraction. Whenever possible the second pre-
molar was the tooth of choice to maximize esthetics (the
extraction site is more visible in lingual treatment) and to
prevent the inset bend of the archwire distal to the canine
from interfering with space closure.
Radiographic assessment
Serial lateral cephalograms were superimposed according
to the method of Pancherz24
(Figure 1). The pre- and post-
treatment difference of the distal contact point of the max-
illary first molar to a line perpendicular to the occlusal
plane through sella (OLp) corresponds to the amount of the
mesial (1) or distal (2) movement of the maxillary first
molar. This includes maxillary growth that occurred in G4.
In this group, the net AL was calculated by subtracting an
average annual amount of maxillary horizontal growth (0.5
mm/year for the females and 0.9 mm/year for the
males).25,26
Dental cast assessment
Dental casts were measured according to the method de-
scribed by Ziegler and Ingervall6
(Figure 2). The posterior
ruga point and the mesial contact point of the first molar
were demarcated on the right and left quadrants, and the
midpalatal raphe was used to construct a median reference
line. The referenced casts were photocopied at 200% en-
largement. The distance (d) between the projections of the
two points to the median reference line was measured for
the right (dR) and left (dL) quadrants and averaged (Figure
2). The difference between pretreatment ‘‘d’’ and posttreat-
ment ‘‘d’’ was defined as AL. Overjet was measured from
the upper and lower dental casts in occlusion. Crowding
was measured as the amount of overlap between teeth.27
Statistics
An analysis of variance with repeated measures was per-
formed to determine statistically significant (P , .05) dif-
ferences between the examined groups for three AL factors
(extraction site, mechanics, and age). Pearson’s correlation
tests examined whether initial crowding and overjet corre-
lated with the amount of AL.
RESULTS
The sex of the individuals was excluded as an AL factor
because no dimorphism was found (P 5 .397).

4. 733
ANCHORAGE LOSS—A MULTIFACTORIAL RESPONSE
Angle Orthodontist, Vol 73, No 6, 2003
FIGURE 2. (a) Pretreatment dental cast; the distance d was defined as the length between the projection of mesial contact point of the first
molar and the projection of the most medial point of the posterior ruga along the midpalatal line. (b) The distance d of the posttreatment dental
cast is shorter because of mesial displacement of the first molar and distal migration of the palatal rugae.
TABLE 1. AL for First (G1) Versus Second (G2) Premolar Extraction
Group n Age (y)
OLp-M1
Before (mm)
OLp-M1
After (mm)
OLp-M1
Difference (mm)
Cephalogram
G1
G2
P
13
13
24.8 6 5.7
24.4 6 5.9
0.874
49.6 6 9.4
45.7 6 9.6
0.299
51.4 6 9.4
48.0 6 9.9
0.372
1.8 6 1.4
2.4 6 1.3
0.332
Dental cast
G1
G2
P
13
10
7.2 6 2.9
6.9 6 2.0
0.771
4.8 6 2.4
4.0 6 2.3
0.404
2.4 6 1.9
2.9 6 1.6
0.332
Extraction site—first (G1) vs second (G2)
premolar extractions
AL, as measured from the cephalometric radiographs,
was 1.8 6 1.4 mm for G1 and 2.4 6 1.3 mm for G2. AL,
measured from the dental casts, was 2.4 6 1.9 mm for G1
and 2.9 6 1.6 mm for G2 (Table 1).
Thus, AL with second premolar extractions was 0.5 6
1.3 mm greater as measured from the cephalometric radio-
graphs and 0.5 6 1.7 mm greater when measured from the
dental casts. The other two AL factors tested, ie, mechanics
(lingual appliances) and age (nongrowing) were similar in
these groups.
Mechanics—lingual (G1) vs labial (G3)
edgewise appliances
AL, when measured from the cephalometric radiographs,
was 1.8 6 1.4 mm for G1 and 3.0 6 1.4 mm for G3. AL,
when measured from dental casts, was 2.4 6 1.9 mm for
G1 and 3.9 6 2.7 mm for G3 (Table 2).
AL was significantly greater in the labial appliance group
than in the lingual group when measured from cephalo-
metric radiographs (P , .05) but was not significantly
greater when measured from dental casts. The other two
AL factors tested, ie, age (nongrowing) and extraction site
(upper first premolar) were similar in these groups.
Age—nongrowing (G3) vs growing (G4) subjects
The AL, as measured from cephalometric radiographs,
was 3.0 6 1.4 mm for G3 and 3.5 6 1.6 mm for G4. AL,
as measured from dental casts, was 3.9 6 2.3 mm for G3
and 4.1 6 2.3 mm for G4 (Table 3).
Thus, the G4 growing patients showed a greater but not
significantly greater AL than the G3 nongrowing patients
with cephalometric measurements and with dental casts.

5. 734 GERON, SHPACK, KANDOS, DAVIDOVITCH, VARDIMON
Angle Orthodontist, Vol 73, No 6, 2003
TABLE 2. AL for Lingual (G1) Versus Labial (G3) Edgewise Appliance
Group n Age (y)
OLp-M1
Before (mm)
OLp-M1
After (mm)
OLp-M1
Difference (mm)
Cephalogram
G1
G3
P
13
15
24.8 6 5.7
20.0 6 5.4
0.25
49.6 6 9.4
49.2 6 12.5
0.923
51.4 6 9.4
52.3 6 12.7
0.865
1.84 6 1.4
3.0 6 1.4
0.042*
Dental cast
G1
G3
P
13
11
7.2 6 2.9
9.2 6 2.8
0.097
4.8 6 2.4
5.3 6 3.9
0.696
2.4 6 1.9
3.9 6 2.7
0.084
TABLE 3. AL for Growing (G4) Versus Non-growing (G3) Subjects
Group n Age (y)
OLp-M1
Before (mm)
OLp-M1
After (mm)
OLp-M1
Difference (mm)
Cephalogram
G4
G3
P
33
15
12.7 6 2.0
20.0 6 5.4
0.0001
48.3 6 10.0
49.2 6 12.5
0.793
51.8 6 10.2
52.2 6 12.7
0.919
3.5 6 1.6
3.0 6 1.4
0.275
Dental cast
G4
G3
P
19
11
9.2 6 3.5
9.2 6 2.8
0.994
5.1 6 3.1
5.3 6 3.9
0.869
4.1 6 2.3
3.9 6 2.3
0.802
TABLE 4. Comparison of Overjet and Crowding in Relation to the Three AL Factors
Site of Extraction
G1 G2 P
Mechanics
G1 G3 P
Age
G4 G3 P
Overjet (mm)
Crowding (mm)
4.1 6 2.7
7.4 6 1.8
5.0 6 2.8
6.6 6 1.6
0.832
0.743
4.1 6 2.7
7.4 6 1.8
5.5 6 2.6
6.7 6 4.2
0.948
0.776
5.4 6 2.5
6.2 6 3.7
5.5 6 2.3
6.7 6 2.0
0.995
0.876
FIGURE 3. An inverse correlation developed between initial crowd-
ing and anchorage loss measured from cephalometric radiographs.
However, after accounting for the maxillary growth expe-
rienced by G4, the net mean AL of G4 was less than that
of G3, with no significant net difference. The other two AL
factors tested, ie, mechanics (labial edgewise appliances)
and extraction site (first premolar) were similar in these two
groups.
Overjet and crowding
Overjet and crowding were compared using ANOVA be-
tween the three-paired groups of the previously examined
AL factors. These comparisons were not significant (Table
4). Pearson’s correlation test was performed between initial
crowding or overjet and AL for the whole sample. A sig-
nificant inverse correlation was found between initial
crowding and AL for both cephalometric and dental cast
measurements (Figure 3; Table 5). The correlation was par-
ticularly high for the cephalometric data (r 5 20.66, P 5
.001). Crowding and overjet were significantly but indi-
rectly and weakly correlated (r 5 20.28, P 5 .009) (Table
5).

6. 735
ANCHORAGE LOSS—A MULTIFACTORIAL RESPONSE
Angle Orthodontist, Vol 73, No 6, 2003
TABLE 5. Correlation for Initial Crowding and Overjet Versus AL, and Crowding Versus Overjet
n
Anchorage Loss
Cephalogram n
Anchorage Loss
Dental Cast n Crowding
Crowding 75 r 5 20.66, P 5 .001 53 r 5 20.31, P 5 .026
Overjet 75 r 5 0.27, P 5 .021 53 r 5 20.11, P 5 .449 90 r 5 20.28, P 5 .009
DISCUSSION
The extent of AL was consistently greater when mea-
sured from dental casts than from cephalometric radio-
graphs by 0.5–0.6 mm, probably because of posterior dis-
placement of the rugae as the anterior teeth were retracted
(Figure 2). Thus, posttreatment d (Figure 2b) decreased be-
cause of molar mesial movement and posterior displace-
ment of the rugae. Consequently, pretreatment (Figure 2a)
minus posttreatment d (Figure 2b) increased. Thus, dental
cast measurements do not reflect pure molar mesial dis-
placement. In addition, in selecting the most anterior ruga
as a reference point, the extent of its posterior displacement
was greater than that found for the most posterior ruga,
probably because of its proximity to the retracted incisors.
These results conflict with studies claiming that the palatal
rugae are stable landmarks28,29
but are in agreement with
other studies that report similar findings.30,31
The cephalo-
metric method is more reliable because the reference point
(OLp) is located posterior to the examined molar.
Extraction site—first vs second
premolar extraction
The long-held dogma that greater AL occurs when sec-
ond rather than first premolars are extracted14,15,17,32
was
weakly supported by the present study. AL was only 0.5
mm greater with second premolar extractions when as-
sessed either from cephalometric radiographs or from den-
tal casts. However, this was not significant, suggesting that
the location of the premolar extraction site cannot be con-
sidered a major AL factor. Statistically, the nonsignificant
difference between the two groups could be related to the
relative small sample size in this comparison, although the
high P value suggests a reliable nonsignificant AL differ-
ence between the first and second premolar extraction op-
tions. Nevertheless, the present finding is supported by Ong
and Woods,16
who reported a 0.2-mm difference. Williams
and Hosila14
compared extraction sites mainly in the lower
jaw and reported a difference in AL in second vs first pre-
molar extractions (1.0 mm), which was twice that found in
the present study (0.5 mm).
More molar mesial movement in both first and second
premolar extractions was reported by Saelens and De Smit17
(4.4 and 5.3 mm, respectively) compared with the present
study (1.8 and 2.4 mm, respectively). This could be related
to the use of lingual appliances that contributed to a reduc-
tion in molar mesial migration. However, because mechan-
ics was a nonvariable parameter, the inference that the ex-
traction site is a secondary AL factor is probably a valid
deduction with regard to maxillary premolars.
Mechanics—lingual vs labial
edgewise appliances
The choice of a labial or lingual edgewise appliance
demonstrated the greatest AL difference of all two-group
comparisons. Cephalometric comparisons (AL 5 1.2 6 2.1
mm, P , .05) suggest that this factor plays a more clini-
cally relevant role in controlling AL than the extraction site
or patient age. The finding that the lingual edgewise tech-
nique demonstrated lower AL suggests that this factor
should be studied more closely. The theory of Alexander
et al12
that disocclusion of the posterior teeth due to the bite
plane built into the maxillary incisor brackets of the lingual
technique decreases resistance to AL is not supported by
the present findings. Engagement of second molars in the
lingual archwire probably enhanced anchorage, which sup-
ports the theory of Storey relating anchorage resistance to
root surface areas. Freeman,33
who calculated this resis-
tance, found that the ‘‘anchorage value’’ (AV) of the max-
illary six anterior teeth (AVa 5 1412) almost equaled that
of the posterior teeth (AVp 5 1574) when the second mo-
lars were excluded but was almost half when the second
molars were included (AVp 5 2474).2,33
However, in G3
subjects, wearing a headgear was expected to offer greater
resistance than inclusion of second molars; but poor patient
compliance related to age (20 6 5.43 years) could explain
some of the greater AL displayed by these subjects.
Kurz and Bennett13
suggest that the smaller arch perim-
eter increases the rigidity of lingual archwires, which may
increase anchorage control during retraction. The larger slot
size of the posterior lingual attachments provides an almost
frictionless sliding retraction with no energy burning.21,34
Takemoto36
suggests that the anchorage value of posterior
teeth in the lingual technique is higher than that in the labial
technique because of the nearness of the lingual brackets
to the center of tooth resistance. Additionally, the direction
of forces during space closure with lingual appliances cre-
ates a buccal root torque and distal rotation of the molar
crown, which produces cortical bone anchorage.
The average traction force was 75 and 50 g per anterior
tooth, in the labial and lingual techniques, respectively. Ad-
ditionally, because the lingual retraction force is generated
by elastomeric chains, their force decay is more than 50%
in four weeks.36,37
This means that the average force applied
during the six-week period (interval between lingual ses-

7. 736 GERON, SHPACK, KANDOS, DAVIDOVITCH, VARDIMON
Angle Orthodontist, Vol 73, No 6, 2003
sions) was about 25 g per tooth. This suggests that, on
average, during appointment intervals, the retraction force
was threefold lower in the lingual than in the labial tech-
nique.
The combination of bidimensional orthodontics with its
inherent less friction during sliding mechanics, with light
orthodontic forces for space closure, the inclusion of the
second molar in the anchorage unit, and the placement of
exaggerated curves in both archwires, together appears to
be a very efficient method for anchorage preservation. This
seems to hold true even in severe Class II cases, regardless
of whether first or second premolars are chosen for extrac-
tion.
Age—growing vs nongrowing patients
Greater AL was found in the adult group when postero-
anterior maxillary growth was compared with the adoles-
cent group. However, the difference between the groups
(0.7 6 1.4 mm) was not significant, which suggests that
age, as an AL factor, was secondary to choice of appliance.
Nevertheless, a significantly greater AL was found in the
adult group of the present study (3.0 6 1.4 mm) compared
with the findings of Harris et al,19
who reported only a
minor AL (0.2 mm), which suggests that this factor merits
further study. It should be mentioned that in the latter
study,19
this difference is explained by the fact that the adult
subjects wore Class II elastics for a longer time. In the
present study, both age groups wore elastics for similar pe-
riods. Thus, the results of Harris et al19
confirm that choice
of appliance is a principal AL factor.
Overjet and crowding
The significant indirect correlation found between initial
crowding and cephalometric AL and dental cast AL sug-
gests that crowding is another principal AL factor. The
present study suggests that the greater the crowding the
lower the AL, which contradicts the popular tenet that the
more the arch length deficiency the greater the anchorage
requirement.1,2
Maxillary arch length deficiency can be expressed as
crowding or overjet. According to biomechanical princi-
ples, less anchorage is required to unravel crowding than
to reduce overjet. Ong and Woods16
support the interpre-
tation of the present study because, on account of less
crowding, they reported greater AL compared with Lup-
panapornlarp and Johnston.18
The finding that AL was weakly correlated to overjet also
contradicts the above tenet. However, the same overjet can
be due to upper incisor proclination or a Class II skeletal
condition. The potential for AL is profoundly greater in the
latter than in the former, which explains why overjet per se
was weakly correlated to AL.
The difference in AL between labial and lingual groups
was 1.2 mm. The difference in AL between a moderately
crowded (;4 mm) and a severely crowded (;10 mm) sub-
ject was two mm (Figure 3). This suggests that crowding
was more relevant than mechanics as the key AL factor.
CONCLUSIONS
The hypothesis that AL is a multifactorial response was
supported by the present study. Although only five AL fac-
tors were examined, a regulatory reaction was found. Pri-
mary AL factors (crowding and mechanics) affected AL
more significantly than did secondary AL factors (age, ex-
traction site, and overjet). A pattern of influence was found,
where crowding (inverted influence, ie, the greater the arch
length deficiency, the lower the AL) was superior to me-
chanics (primary AL factors), and extraction site was more
influential than age and overjet (secondary AL factors).
The desire to minimize AL is of major concern because
residual overjet, noncusp fossa relationships, and deep bite
are affected. The study suggests that incorporation of the
second molars in the anchorage strategy, low retraction
forces, and frictionless mechanics are superior to the con-
ventional anchorage means such as headgear or non en
masse retraction.
ACKNOWLEDGMENTS
The authors thank Ms Ilana Gelerenter TAU Statistic Center for
statistical analysis, Ms Anna Bahar for graphical help, and Ms Rita
Lazar for editorial assistance.
REFERENCES
1. Lindquist JT. The edgewise appliance. In: Graber TM, ed. Ortho-
dontics: Current Principles and Techniques. St Louis, Miss: Mos-
by; 1985:565–640.
2. Proffit WR. Biomechanics and mechanics. In: Proffit WR, Fields
HW Jr, eds. Contemporary Orthodontics. St Louis, Miss: Mosby;
2000:295–362.
3. Nanda R, Kuhlberg A. Biomechanical basis of extraction closure.
In: Nanda R, Kuhlberg A, eds. Biomechanics in Clinical Ortho-
dontics. Philadelphia, Pa: WB Saunders; 1996:156–187.
4. Hart A, Taft L, Greenberg SN. The effectiveness of differential
moments in establishing and maintaining anchorage. Am J Orthod
Dentofacial Orthop. 1992;102:434–442.
5. Rajcich MM, Sadowsky C. Efficacy of intra-arch mechanics using
differential moments for achieving anchorage control in extrac-
tion cases. Am J Orthod Dentofacial Orthop. 1997;112:441–448.
6. Ziegler P, Ingervall B. A clinical study of maxillary canine re-
traction with a retraction spring and with sliding mechanics. Am
J Orthod Dentofacial Orthop. 1989;95:99–106.
7. Lotzof LP, Fine HA. Canine retraction: a comparison of two
preadjusted bracket systems. Am J Orthod Dentofacial Orthop.
1996;110:191–196.
8. Gianelly AA. Distal movement of the maxillary molars. Am J
Orthod Dentofacial Orthop. 1998;114:66–72.
9. Bondemark L, Kurol J. Class II correction with magnets and su-
perelastic coils followed by straight-wire mechanotherapy. Occlu-
sal changes during and after dental therapy. J Orofac Orthop.
1998;59:27–38.
10. Byloff FK, Darendeliler MA. Distal molar movement using the
pendulum appliance. Part 1. Clinical and radiological evaluation.
Angle Orthod. 1997;67:249–260.

8. 737
ANCHORAGE LOSS—A MULTIFACTORIAL RESPONSE
Angle Orthodontist, Vol 73, No 6, 2003
11. Wehrbein H, Feifel H, Diedrich P. Palatal implant anchorage re-
inforcement of posterior teeth: a prospective study. Am J Orthod
Dentofacial Orthop. 1999;116:678–686.
12. Alexander CM, Alexander RG, Gorman JC, et al. Lingual ortho-
dontics: a status report. Part 5. Lingual mechanotherapy. J Clin
Orthod. 1983;17:99–115.
13. Kurz C, Bennett R. Extraction cases and the lingual appliance. J
Am Ling Orthod Assoc. 1988;3:10–13.
14. Williams R, Hosila FJ. The effect of different extraction sites
upon incisor retraction. Am J Orthod. 1976;69:388–410.
15. Creekmore TD. Where teeth should be positioned in the face and
jaws and how to get them there. J Clin Orthod. 1997;31:586–
608.
16. Ong HB, Woods MG. An occlusal and cephalometric analysis of
maxillary first and second premolar extraction effects. Angle Or-
thod. 2001;71:90–102.
17. Saelens NA, De Smit AA. Therapeutic changes in extraction ver-
sus non-extraction orthodontic treatment. Eur J Orthod. 1998;20:
225–236.
18. Luppanapornlarp S, Johnston LE Jr. The effects of premolar-ex-
traction: a long-term comparison of outcomes in ‘‘clear-cut’’ ex-
traction and non-extraction Class II patients. Angle Orthod. 1993;
63:257–272.
19. Harris EF, Dyer GS, Vaden JL. Age effects on orthodontic treat-
ment: skeletodental assessment from Johnston analysis. Am J Or-
thod Dentofacial Orthop. 1991;100:531–536.
20. Bennet JC, McLaughlin RP. Controlled space closure with a
preadjusted appliance system. J Clin Orthod. 1990;24;4:251–260.
21. Gianelly AA, Bednar JR, Dietz VS. A bidimensional edgewise
technique. J Clin Orthod. 1985;19:418–421.
22. Geron S. The lingual bracket jig. J Clin Orthod. 1999;32:457–
462.
23. Pancherz H. The mechanism of Class II correction in Herbst ap-
pliance treatment. A cephalometric investigation. Am J Orthod.
1982;82:104–113.
24. Pancherz H. A cephalometric analysis of skeletal and dental
changes contributing to Class II correction in activator treatment.
Am J Orthod. 1984;85:125–134.
25. Bishara SE, Peterson LC, Bishara EC. Changes in facial dimen-
sions and relationships between the ages of 5 and 25 years. Am
J Orthod Dentofacial Orthop. 1984;85:238–252.
26. Sinclair PM, Little RM. Dentofacial maturation of untreated nor-
mals. Am J Orthod. 1985;88:146–156.
27. Van der Linden FP. Changes in the position of posterior teeth in
relation to ruga points. Am J Orthod. 1978;74:142–161.
28. Haruki T, Little RM. Early versus late treatment of crowded first
premolar extraction cases: postretention evaluation of stability
and relapse. Angle Orthod. 1998;68:61–68.
29. Hoggan BR, Sadowsky C. The use of the palatal rugae for the
assessment of anteroposterior tooth movements. Am J Orthod
Dentofacial Orthop. 2001;119:482–488.
30. Almeida MA, Philips C, Kula K, Tulloch C. Stability of the pal-
atal rugae as landmarks for analysis of dental casts. Angle Orthod.
1995;65:43–48.
31. Bailey LT, Esmailnejad A, Almeida MA. Stability of the palatal
rugae as landmarks for analysis of dental casts in extraction and
nonextraction cases. Angle Orthod. 1996;66:73–78.
32. Schwab DT. The borderline patient and tooth removal. Am J Or-
thod. 1971;59:126–145.
33. Freeman DC. Root Surface Area Related to Anchorage in the
Begg Technique [MS thesis]. Memphis, Tenn: University of Ten-
nessee. 1995.
34. Schudy GF. The Bimetric system. Am J Orthod. 1975;67:57–91.
35. Takemoto K. Anchorage control in lingual orthodontics. In: Ro-
mano R, ed. Lingual Orthodontics. Hamilton, Canada: BC Deck-
er; 1998:75–82.
36. Stevenson JS, Kusy RP. Force application and decay character-
istics of untreated and treated polyurethane elastomeric chains.
Angle Orthod. 1994;64:455–464.
37. Wong AK. Orthodontic elastic materials. Angle Orthod. 1976;46:
195–205.